The Twin Delayed Deep Deterministic Policy Gradients Algorithm or TD3 for short, is an extension of traditional Deep Q-Learning which is religated to a discrete action space.
The TD3 algorithm was conceived by the research team at Google Deep Mind (here)
In this project, I will be implementing the TD3 algorithm using PyTorch to train a Quadruped to run across a field generated in the pybullet enviornment.